A Revised Classification of Cynodonts (Reptilia; Therapsida)
Total Page:16
File Type:pdf, Size:1020Kb
71 Palaeont. afr., 14.71-85.1972 A REVISED CLASSIFICATION OF CYNODONTS (REPTILIA; THERAPSIDA) by * James A. Hopson and tJames W. Kitching INTRODUCTION We wish to thank the following colleagues for Cynodonts are very advanced mammal-like access to unpublished information which has been incorpo~ated ~nto th~s ~eptiles of the Pe:m<;>-Triassic which are of special paper: Dr. J. F. Bonaparte, Interest to evolutIOnIsts because they gave rise to FundacIOn MIgUel LIllo, Tucuman, Argentina; Dr. t~e Class Mammalia during Middle or Late Triassic A. W. Crompton, Museum of Comparative tIme. Cynodonts have been known from strata of Zoology, Har~' ard University; and Mr. J. W. A. van Early Triassic age in South Africa for over one Heerden, NatIOnal Museum, Bloemfontein. hundred years, and numerous specimens have been Several aspect~ .of this classification require collected and described. In recent years the record comment. In deCIdIng whether to consider a of cynodonts has been extended into earlier and generic or specific name to be valid, we have taken later time zones, not only in southern Africa but the ~osition that. the burden of proof is on the in East Africa, South America, Russia , China', and , descnber to conVInce us that the named taxon is most recently, in North America. Much of the distinct from earlier-named taxa. Where there is a material from outside of Africa has not yet been reasonably high probability that two named taxa fully described. are synonymous, we have usually synonymized . Ap'prox~m~tely 125 species of cynodonts them, even though the evidence for their identity is (Includmg IctIdosaurs and tritylodonts herein not conclusive. We have done this because: (1) the considered to be cynodonts) have been n~med. Of cynodonts have been excessively split (as, in fact, these, however, only a very small handful has been ha~e . most therapsid groups), so that our first adequately characterized so that anyone species pnonty seemed to be the reduction of the can be reliably distinguished from all others. As a confusing welter of inadequately-characterized result of the taxonomic confusion which prevails in gen~ra and species to fewer, more adequately defIned taxa; (2) many of the type specimens are the Cy:nodo~ti~, our knowledge of the patterns of evolutIOn wIthm the group is based on detailed f:agmentary and poorly-preserved so that diagnos knowledge of a few species. tIC ~harac~ers are lackIng; (3) where locality and . The stimulus for the present revised classifica stratIgraphIc data are available, it is evident that an tIOn of cynodonts was an extended research visit to excessive number of closely-related species have the Bernard Price Institute for Palaeontological been na!ll~d from very limited geographic areas and Research by the senior author in late 1971. The from wIthIn very narrow stratigraphic intervals. In purpose of the visit was to study primitive those cases where taxa based on inadequate types cy?odonts, but all available cynodont material in could be reasonably synonymized with adequately thIS and many other South African museums was characterized species, we did so but where examined. tt The junior author has also studied most inadequate types could not be so allo~ated we have of th.e .cynodont mate:ial in South Africa as part of chosen to declare the names on which they are a reVISIOn of the stratIgraphy and vertebrate fauna based to be nomina vana. of · the Beaufort Series. Many of the taxonomic . The highe~-level classification adopted here dIffers substantIally from that used in most earlier conclusions independently reached by the two classifications (e.g., Watson and Romer, 1956; ' ~uthors have proven to be identical, though based In great part on different evidence. Because of the Romer, 1966; Lehman, 1961; but see Haughton ~nd Bri~k, 1954). We have abandoned separate larg~ amount of overlap in our work, we have decIded to publish our taxonomic conclusions as a In~raordInal . rank for the "Ictidosauria" (i.e., TntheledontIdae) and Tritylodontidae; instead joint report, ~eserving for separate future papers the presentatIOn of the detailed evidence upon they are listed as families within the Infraorder which these conclusions are based. For the sake of Cy~od(:mtia. These groups are clearly cynodont completeness, the senior author has reviewed the denvatIves and we believe that their positions as non-Sout~ A.frican cynodonts, though many of the * James A. Hopson, The University of Chicago, Department of taxonomIC Judgments on these forms are less Anatomy, 1025 East 57th Street, Chicago, Illinois 60637, U.S.A. securely b~sed than are the judgments on the t Jam.es W. Kitching, University of the Witwatersrand, Bernard Price South Afncan cynodonts. The revision of the InstItute for Palaeontological Research, Milner Park, Johannes burg, South Africa. family Tritylodontidae is based on the senior tt This research was supported by a grant from the National author's unpublished studies. Science Foundation. 72 advanced end-members of two major adaptive lines root genera, Cynognathus and Tritylodon and (2) within the cynodont radiation are best indicated the fact that these genera represent extremes of by incorporating them within the Cynodontia. specialization within their respective groups. The Their respective positions as the culmination of the resulting classification, we believe, possesses the carnivorous and herbivorous rami of the cynodonts advantage of expressing both phyletic relationships has been expressed in the classification by dividing and adaptive trends better than those classifica the infraorder into two main groups at the level of tions which place the cynodonts and their superfamily. For the carnivorous group, we have therapsid descendants in three infraorders (Cyno adopted Brink's (1963) term Cynognathoidea; for dontia, Ictidosauria, and Tritylodontia). A similar the herbivorous group, we have chosen to use sort of reclassification of other therapsid groups, Simpson's (1928) term Tritylodontoidea. These notably of the therocephalian-bauriamorph choices were determined by: (1) familiarity of the assemblage, is also overdue. INFRAORDER CYNODONTIA SUPERFAMILY CYNOGNATHOIDEA Brink 1963. Cynodonts in which the posterior postcanine teeth usually possess three or more cusps aligned in an anteroposterior row; a narrow lingual cingulum bearing several small cusps is frequently present, but only rarely is the cingulum expanded to form a broad lingual shelf; canines are usually well-developed; incisors are usually small and unspecialized. Most members are carnivorous. Included are primitive cynodonts of the Permian and earliest Triassic and all carnivorous cynodonts of the later Triassic. FAMILY PROCYNOSUCHIDAE Broom 1948. Includes: Silphedestidae Haughton and Brink 1954; Dviniidae Tatarinov 1968 a. Primitive cynodonts with five or more incisors, small "precanine" maxillary' teeth in front of the enlarged canine, and posterior postcanine teeth with a prominent lingual cingulum; the dentary is small, with a low coronoid process, no angle, and with the masseteric fossa restricted to the posterodorsal portion of the bone; palatal plates of the maxillae and palatines do not meet at the midline so that the secondary palate is incomplete; interpterygoidal vacuities usually present; ribs normal, lacking plate-like expansions. Horizon: Upper Permian: Daptocephalus Zone, Beaufort Series, of South Africa; Zone IV of Russia; Kawinga Formation of Tanzania. GenusPROCYNOSUCHUS Broom 1937. Synonyms: ? Cyrbasiodon Broom 1931; Paracynosuchus Broom 1940a; Mygalesaurus Broom 1942; Aelurodraco Broom and Robinson 1948b; Leavachia Broom 1948; Galeophrys Broom 1948; Galecranium Broom 1948; Silphedestes Broom 1949; Protocynodon Broom 1949; Silphedocynodon Brink 1951; Scalopocynodon Brink 1961. PROCYNOSUCHUS DELAHARPEAE Broom 1937. Synonyms: ? Cyrbasiodon boycei Broom 1931; Procynosuchus rubidgei Broom 1938; Paracynosuchus rubidgei Broom 1940a; Nanictosuchus melinodon Broom 1940a; Nanicto saurus robustus Broom 1940b; Mygalesaurus platyceps Broom 1942; Aelurodraco microps Broom and Robinson 1948b; Leavachia duvenhagei Broom 1948; Galeophrys kitchingi Broom 1948; Galecranium liorhynchus Broom 1948; Silphedestes polyodon Broom 1949; Protocynodon pricei Broom 1949; Silphedocynodon gymnotemporalis Brink 1951; Leavachia microps Brink and Kitching 1951a; Leavachia gracilis Brink and Kitching 1951a; Scalopo cynodon gracilis Brink 1961. Horizon: Upper Permian: Daptocephalus Zone of South Africa. Remarks: Many genera and species have been based on juvenile specimens of Procynosuchus, including all forms placed in the Family Silphedestidae by Haughton and Brink (1954). Cyrbasiodon boycei is probably synonymous with P. 73 delaharpeae, but in view of the poor quality of the type specimen of the former species, we have chosen to retain the latter well-established name. (See paper by Mendrez in this volume, p. 51). Genus D VINIA Amalitzky 1922. Synonyms: Permocynodon Woodward 1932. D VINIA PRIMA Amalitzky 1922. Synonyms: Permocynodon sushkini Woodward 1932. Horizon: Upper Permian: Upper Tatarian deposits of Arkhangel'sk region, Russia. Genus PARATHRINAXODON Parrington 1936. PARATHRINAXODON PROOPS Parrington 1936. Horizon: Upper Permian: Kawinga Formation of the Ruhuhu Valley, Tanzania. PROCYNOSUCHIDAE incertae sedis: Genus NANOCYNODON Tatarinov 1968b. NANOCYNODON SEDUCTUS Tatarinov 1968b. Horizon: Upper Permian: Upper Tatarian of Russia. Remarks: Tatarinov placed this species in the Galesauridae, but it is more likely a juvenile